PRNet: Self-Supervised Learning for Partial-to-Partial Registration

TL;DR

PRNet introduces a self-supervised deep learning framework for partial-to-partial point cloud registration, outperforming existing methods and enabling transferable representations for classification tasks.

Contribution

It is the first to propose a self-supervised, deep learning-based approach specifically designed for partial-to-partial registration, jointly learning keypoints and correspondences.

Findings

Outperforms PointNetLK, DCP, and non-learning methods on synthetic data.

Learns transferable geometric representations for classification.

Predicts consistent keypoints and correspondences across views and objects.

Abstract

We present a simple, flexible, and general framework titled Partial Registration Network (PRNet), for partial-to-partial point cloud registration. Inspired by recently-proposed learning-based methods for registration, we use deep networks to tackle non-convexity of the alignment and partial correspondence problems. While previous learning-based methods assume the entire shape is visible, PRNet is suitable for partial-to-partial registration, outperforming PointNetLK, DCP, and non-learning methods on synthetic data. PRNet is self-supervised, jointly learning an appropriate geometric representation, a keypoint detector that finds points in common between partial views, and keypoint-to-keypoint correspondences. We show PRNet predicts keypoints and correspondences consistently across views and objects. Furthermore, the learned representation is transferable to classification.